Run the TS E0 check for reactions with a monoatomic participant

[calvinp0]

The E0 sanity check for a transition state was gated on every reaction participant having both a converged single-point energy and frequencies.

A monoatomic participant — [O], [H], [Cl] — never runs a frequency job, because it has no vibrational modes. So that gate could never be satisfied, and the E0 check silently never ran for any reaction involving an atom.

The fix, in two halves

Readiness. A new species_is_ready_for_e0() requires an SP energy from all participants but frequencies only from the non-monoatomic ones. Both readiness sites route through it.

The trigger. post_sp_actions() called check_rxn_e0_by_spc() only when species_has_freq() was true. So when the last job to complete was an atom's single-point, the newly-satisfied reaction was never re-evaluated, and the TS kept an unvalidated ts_checks['E0'] is None for the rest of the run. It now re-checks on the same readiness predicate.

The participant loop also iterates set(labels) and indexes self.output directly, rather than filtering self.output.items() — so a participant absent from self.output now fails the gate instead of being skipped as vacuously ready.

Tests

Three, covering the predicate itself; a monoatomic participant reaching check_ts and driving switch_ts on a failed E0; and an atomic SP completing last as the event that triggers the check.

None of them touch the class-level sched1/sched3 fixtures — two build object.__new__(Scheduler) directly and one calls the module function — so they are unaffected by #1009's setUpClass → setUp restructuring of TestScheduler.

Verification

Rebased from 146 behind onto current main. One conflict, in arc/scheduler_test.py's from arc.scheduler import (...) block: main had added tsg_method_matches_adapter to the same line this branch extended with species_is_ready_for_e0. Resolved as a union across two lines.

Content identity against the old merge-base accounts for every difference — all four changed lines are that one resolution, and there is no unexplained only-in-old line, so no hunk was dropped. arc/scheduler.py auto-merged; check_rxn_e0_by_spc, post_sp_actions and the new helper were each re-read on the new base, and ARCSpecies.is_monoatomic() still exists on main.

Tests serial: 82 passed on the pre-rebase head → 99 passed after (main added tests). Nothing introduced. One commit, two files, each touched once.

Note

The original commit message is a single subject line with no body, so it records no measurement or failure-mode detail beyond what the diff shows. The description above is derived from the code.

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 64.62%. Comparing base (45d73a0) to head (8bafd1e).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1012      +/-   ##
==========================================
+ Coverage   64.60%   64.62%   +0.02%     
==========================================
  Files         119      119              
  Lines       39785    39789       +4     
  Branches    10307    10308       +1     
==========================================
+ Hits        25703    25714      +11     
+ Misses      11105    11096       -9     
- Partials     2977     2979       +2     

Flag	Coverage Δ
functionaltests	64.62% <ø> (+0.02%)	⬆️
unittests	64.62% <ø> (+0.02%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.